Page MenuHome

Irradiance volume Distance should not be affected by scale
Open, NormalPublic


System Information
Operating system: Windows-10-10.0.17134 64 Bits
Graphics card: GeForce GTX 970/PCIe/SSE2 NVIDIA Corporation 4.5.0 NVIDIA 416.34

Blender Version
Broken: version: 2.80 (sub 51), branch: blender2.7, commit date: 2019-03-25 00:55, hash: rB07f6be87a97e
Worked: N/A

Short description of error
When you scale the Irradiance Volume to the size of a room (e.g. 6m x 4m x 2.4m), then the Distance is scaled too, and hence scaled non proportionally.

Furthermore I expected the Distance to represent an actual Distance, in particular because it expressed as one (e.g. 1 metre).

Exact steps for others to reproduce the error

  1. Create a Cube which will be sized to be a room. Set Dimensions to X=6m, Y=4m, Z=2.4m. (This is just to compare against the irradiance volume.)
  2. Create an Irradiance Volume. Scale to X=2, Y=1.2, Z=1. Set Irradiance Distance to 1m
  3. Observe that the Irradiance volume now overlaps the Cube, and that due to the non-proportional scaling there is more irradiance influence in the X direction compared to Z.

Why does this matter? I would think that for architectural scenes you probably want to carefully control the irradiance volume influence distance so that indirect lighting does not extend through the wall into an adjoining room. Furthermore you do want to ensure the irradiance volume does correctly overlap the walls, and in some cases it might be important to position the irradiance probes close to the wall. Using the current non-proportional scaling, and the Distance that is actually a factor it is hard to achieve this.

Note if I import a file from 2.79 where the units were set to centimetres then unit scale is set to 0.01. This means that the Irradiance Volume requires massive scaling to get it to the correct size (e.g. 200, 120, 100) - hence the Distance must be set to a very small number - specifically 0.01m.
I note that if I create a file in 2.80 then when setting the Length to Centimetres it does not change the Unit Scale. So perhaps this is now handled better.

Based on the default startup or an attached .blend file (as simple as possible).
I can't see where to do this - however the blend file is not particularly important.



Event Timeline

Mark Stead (Scumbag) updated the task description. (Show Details)
Philipp Oeser (lichtwerk) triaged this task as Normal priority.Tue, Apr 30, 4:22 PM

Not sure if there are reasons to keep it like this, but I see the point here.
If we look at forcefields for example, their power and especially their min and max distance also dont scale with the obmat.

Just a thought... but will put on "Normal" prio for now...
(maybe @Brecht Van Lommel (brecht) also has an opinion?)